Glucocorticoids (G) are used in virtually all modern treatment regimens for childhood acute lymphoblastic leukemia (ALL) even though certain patients' blasts are G resistant at diagnosis and(or) at relapse. Since neither the mechanism(s) nor the means to identify G resistant cells have been defined, some patients receive a drug that for them has no therapeutic benefit. The immediate objective of this proposal is to identify specific defects in leukemic cell glucocorticoid receptors (GR) that cause resistance in childhood ALL. The first mechanism to be investigated is that G resistance is due to defects in GR that alter nuclear translocation and(or) GR activation, an obligatory step preceeding translocation. Leukemic blasts from a large group of uniformly treated children with ALL will be assayed for nuclear translocation of GR by isolating nuclei after incubation with [3H]labeled dexamethasone. Possible alterations in G activation will be assessed by quantitating the binding of [3H]triamcinolone-GR to minicolumns of DNA cellulose, DEAE cellulose, and hydroxylapatite. The consequences of GR defects on G action will be determined in vitro by measuring G-induced inhibition of glucose uptake and the in vivo cytolytic response to a 2 to 3 day single-agent G trial in selected patients. Comparisons will be made between diagnosis and relapse to determine possible therapy-induced changes in GR and responsiveness. The second mechanisms to account for clinical G resistance that we will investigate is that certain blast populations contain both GR-positive sensitive cells and GR-negative resistant cells--the relative proportion of each cell type being related to the degree of G sensitivity. Antibody to purified GR will be used to establish an immunocytochemical assay for GR in cell smears. Comparison with whole cell radioreceptor assays will be used to determine if the wide range in GR levels in patient samples is due to blast heterogeneity, and if therapy changes, the proportion of GR positive cells at relapse. These results will be correlated with in vivo and in vitro G responsiveness. Results from these studies should provide additional information on the biology of leukemia at diagnosis and relapse, define mechanisms for the acquisition of G resistance, and may, ultimately, be useful to define more effective treatment regimens for individual patients with ALL; particularly those at relapse. (D)